19 Feb
2019
19 Feb
'19
4:25 p.m.
On Tue, 19 Feb 2019, dwight wrote:
Actually, I'd like to see it just read/write flux
changes + index marks
onto a SD card for later analysis. Building all the smarts into the
controller means that some formats will get missed. One can later write
translation code for what ever format one has. Make this information
open source, much of it is already in bits and pieces. Make sure it can
read and buffer an entire track of data in RAM ( Gotek can't ).
As I see it, flux transition hardware COULD be all that is needed for
hardware. Emulation of FDC could be done in software with flux transition
hardware.
First stage would be with minimal software, to save the track flux
transitions in a file.
Next stage, If and when a 765 FDC can be emulated.
THEN, make the FDC emulation code loadable into RAM, and repoint the
Int13h vector to it. (TSR)
THEN, make new version that adds TRACK-READ (ala WD, NOT the multi-sector
read in the 765)
Then fine details, such as whether or not to implement (switchable) index
pulse "flash blindness", inability to handle 128 byte sectors, etc.
OR, that much could be done without full flux transition functions.
But, if you DO retain full access to the flux transition data,
THEN, add GCR decoding, and add that into the FDC emulation (so that
Int13h/fn2 could read GCR sectors, etc.)
THEN, add hard sector handling
THEN create IFS (Installable File System), which is totally not connected
with any of this, except when needed for non 765 compatible physical
formats. That is what XenoCopy would have eventually become, IFF I were to
have been able to continue.
(Uniform did implement most of an installable file system for CP/M
formats)
The end-user wants to load their Epson QX-10 documents and Kaypro CP/M
documents into Word. They don't want to know how it is done, nor go
through the essential steps of reading alien sectors, processing alien
file system to select appropriate sectors, AND convert word processor
formatting codes from one word processor to another. Do you think that
you could teach them how? XenoCopy transferred FILES; but left the
content alone. So, YES, I did get people telling me that XenoCopy had
"trashed the last letter of every word in the Wordstar files"
I drew the line between transferring files V modification of the content
of the files. You are drawing the line between flux transition data V
sectors/file system/files/content
--
Grumpy Ol' Fred cisin at xenosoft.com